Information Systems - The State of the Field
.pdf74 Crisis in the IS Field?
years, capturing as many IS specialties as is humanly possible. Fortunately, we can draw upon some of our earlier work for inspiration and guidance. Our prior work has on the one hand been more concrete yet on the other, more narrowly focused. The internal issues of the field that we will raise have, to some extent, been foreshadowed in Hirschheim (1986a), Klein and Hirschheim (1991) and Klein (1999). This second article introduced the dualism between purposive and communicative rationality in the context of IS development methods and tools. This distinction plays a critical and much broader role in this paper. The external issues of the field were touched on in Hirschheim and Klein (2000) where we first started to worry about the field’s disconnects with its external stakeholders. Some of these concerns were also foreshadowed in an earlier paper which looked at societal change and its potential impact on the field (Klein and Hirschheim 1987).
In this paper, we will make a case for the members of the IS field to devote more attention than in the past to the potential role of IS in supporting sincere, reflective, agreement-oriented, yet critical conversations and debates. IS can support the achievement of agreement through the force of rational argumentation in two ways: as a convenient communication medium that reduces the transaction costs of communications; and as potential content provider that gives access to relevant information for evidence giving in critical debates that otherwise would not be available to the participants. This conversa- tional-critical potential of IS was described in Klein and Lyytinen (1992) and then again in Hirschheim et al. (1995, pp. 165–169).
With this in mind, we chose to structure the paper as follows. In section 2, we first offer a preview of the line of argument that the paper takes. This may seem somewhat unusual, but because of the nature and complexity of our argument, some up front summary/ preview will help the reader follow our train of thought. After the preview, we attempt to assess the status of the IS field,4 i.e., from where we have come. We offer a brief reflective history noting that the field is fragmented, and our historical treatment helps to better understand why it is fragmented. In sections 3 and 4, we explore the current state of the field. We note that because of the disparate stakeholders the field tries to satisfy, a number of communication gaps are apparent. We term these ‘disconnects’. These disconnects are discussed in terms of two communities: the external community and the internal community. The external disconnects are associated with (1) IS researcher-executive, (2) IS practitioner-IS researcher, and (3) executiveIS practitioner communication gaps. The internal disconnects relate to communication gaps between (4) IS researcher-IS researcher, and
(5) IS researcher—other disciplinary researcher that significantly
From Where Have We Come 75
influence the well being of IS in the academy. In section 5, we offer a proposal for the direction that corrective action could follow. This is our suggestion for ‘where might we go from here?’ We discuss the different types of knowledge that may be useful in IS and following on from there, the potential value of creating a shared structure which creates a core Body of Knowledge (BoK) out of parts that are now mostly disjoint. In section 6, we derive some implications and then recommendations from our ideas, particularly as they relate to the setting of research agendas, changes in the institutional publication practices, and the development and use of knowledge creation and transformation networks. Last, we speculate on a new frontier for IS revolving around a new cyber democracy.
2 FROM WHERE HAVE WE COME
2.1 Synopsis of Our Argument5
Having been in the field for a combined 60+ years, we have seen considerable growth in and maturation of the field. This includes many fads, but also substantial structural and institutional changes, as well as numerous calls for the need to change research methods, publication practices, tenure and promotion criteria, editorial policy, and so on. However, even with the normal ‘give and take’ and the cycles that are common to any field, we feel particularly uncomfortable with the current state of the IS field because we see certain underlying structural patterns that give us serious cause for concern. In this paper, we attempt to analyze what these structural patterns are and why they have come about. We also articulate what is deeply disturbing about these patterns. Additionally, we offer a knowledge structuring and social communication perspective on assembling and sharing IS knowledge; viz. the concept of an IS Body of Knowledge.6 This paper will proceed from a historical background sketch, to a diagnosis of structural deficiencies in IS as a field, to suggesting some directions for corrective action.
We need to start first by identifying the underlying structural patterns from which our concerns arise by presenting a brief historical reconstruction of the evolution of IS as an academic field. We believe that the field needs to understand where it has come from in order to better discern where it might go.7 This will also shed light on the fundamental assumptions that lie behind our concerns for the viability of IS as a discipline. The goal of this paper then is to reflect upon what has occurred in our discipline over the past thirty plus years and to consider the possible course of events that would favor
76 Crisis in the IS Field?
or jeopardize the long-term survival of the field. This goal suggests two purposes for this paper: a diagnostic one, which is the key focus of the paper, and a therapeutic one where we offer our recommendations on how the field needs to change its strategic research priorities and institutional arrangements.
2.1.1 Diagnostic purpose
First, from a diagnostic perspective, we find that the IS discipline suffers from two problematic structural patterns: (1) a state of fragmentation, and (2) a number of significant communication gaps, which we term ‘disconnects.’ These lead to at least the following major threats: intellectual rigidity and the subsequent lack of fruitful discipline-wide communication; and a lack of relevance leading to possible emasculation through dispersal into other disciplines or business functions or through ‘offshoring.’ If so, the institutionalized IS discipline, as we know it, may cease to exist, at least in most of the Western World. Unless the academic leaders of the field begin to address these structural threats, they will eventually undermine the viability of the field. IS as a discipline could fail internally from a lack of integration, splitting into separate specialties that can then easily be allocated to other management disciplines or be offshored.8 Practitioners are facing a similar threat directly and, if they are absorbed or dispersed, academicians will soon follow, for without a thriving practitioner community, there is little need for an academic one. If so, IS would have failed externally for the lack of service it provided to its principal external clients: IS practitioners, business unit users including senior manage- ment—CEOs, CFOs, COOs, etc. In this paper, we seek to address these structural deficiencies that are not connected to any economic cycle or ‘market crash.’ To meet this goal, we present a problem diagnosis that arises from a historically informed reflection on the current state of our discipline.
2.1.2 Therapeutic purpose
Second, from a therapeutic perspective, we contend that IS as a discipline has both communicative and purposive-rational functions,9 and that IS research and practice straddle the methodological divide between these functions. Both functions are equally important for IS as a whole, but not necessarily for each of its sub-specialties. Because of the important role that the purposive-rational and communicative functions play, we offer a brief discussion of them.
Purposive rationality is exclusively concerned with achieving given ends with the minimum expenditure of means. The purposive-rational function of IS takes the calculative optimization of the means-ends
From Where Have We Come 77
relationship as the guiding principle of human action, as is presumed in mainstream economic theory and the engineering sciences. For example, a database programmer seeking to minimize response times while keeping duplication and storage needs to a minimum acts in a purposive-rational manner. In distinction to the means-ends orientation of purposive-rational action, the communicative function of IS research and practice attempts to contribute to the achievement of mutual understandings or at least compromises between different agents through negotiated arrangements. The analyst trying to negotiate the definition of requirements with different user groups is oriented toward reaching mutual understanding and agreement (unless he chooses to act strategically using deception, cf. Keen’s 1981 description of implementation and counter-implementation games). Therefore, the communicative function of IS needs to examine how mutual understanding and agreements can be achieved in different situations. For this purpose, it is concerned with forming shared interpretations of norms, meanings and values and with maintaining social relationships. The further explanation of the communicative function draws on a general theory of the nature of human communication in everyday life. It analyses how through sincere conversations and discourses on any topic of interest, people can reach mutual understanding through having a minimum of common background assumptions about the world. This also applies to academic discourse in research communities (cf. Heath 2001, p. 17; McCarthy 1982, Ch. 4; Habermas 1984, p. 75). We have elaborated on this in earlier research (e.g., Klein and Hirschheim 1991; Hirschheim, et al. 1996) based on the general framework of Habermas ‘Theory of Communicative Action’ (Habermas 1984, 1987).
It is important to realize that the type of knowledge and its role is very different in the purposive-rational and communicative functions of IS research and practice. The principal role of knowledge in the pursuit of rational-purposive action is to acquire powers of prediction and outcome control. The type of knowledge that is most useful for this is nomological, of which the laws of nature are the most typical example. If purposive-rational strategies are directed against human agents, they involve a claim to social power and treat people either as passive, inanimate objects or as opponents capable of intelligent counterstrategies (for an example of this in IS see Keen, 1981). In contrast, communicative action is based on knowledge of social norms, conventions, habits and historically accepted viewpoints as are typically expressed in ordinary language. Other agents are not treated as inanimate objects or opponents, but as fellow human beings and partners. The kind of knowledge that is most useful in agreement oriented interactions comes from the humanities, especially history, study of
78 Crisis in the IS Field?
foreign languages and cultures, comparative literature and social anthropology Therefore, the communicative function brings IS close to the methods of history, social anthropology, social psychology, sociology and—at the philosophical level—hermeneutics.10
While important strides have been made in both the communicative and purposive-rational functions of IS research and practice, their full extent and the interconnections between them have never been widely recognized nor become integral to our culture as an applied disci- pline—as pointed out in the literature analysis in Hirschheim et al. (1996). The communicative function for IS professionals and researchers first should include a shared understanding of our short history—the major intellectual waves that shaped our perspectives. Most of these intellectual waves originated in Europe, in particular the U.K. and Scandinavia, and the U.S. We believe the first time prominent representatives of the differing research world views of the U.S. and Europe engaged in an intense, face-to-face dialogue, was during the so-called ‘Manchester conference’ (Mumford et al. 1985). Since then, the US/ Europe divide has gradually been broken down. For example, the original Conference on Information Systems (CIS) has become ICIS (the International Conference on Information Systems); AIS—our institutional IS academic body—has a membership consisting of a significant and growing number of international affiliates. Yet, only a few old-timers, who directly participated in the beginnings of the globalization of IS research, know the intellectual foundations that drove these institutional changes and which now legitimize them. Therefore, a historical reflection, biased and incomplete as it necessarily must be here, can provide an essential foundation for a broader dialogue—internally. The second important focus of the communicative function looks outward to our clients. It suggests that all IS researchers need a better understanding of their clients’ ‘lifeworld’ and existential concerns (cf. Habermas 1984). The past examples for conveying such understandings were important, but too limited. They consisted of yearly surveys of ‘Key Issues Facing CIOs,’ SIM APC grants, and a few select conferences. Some of the most successful ones were the IS research centers at various universities. However, their reach was rather limited as they served only a handful of elite universities that enjoyed the location and resources to afford them.
2.1.3 A proposed solution—a consensual body of knowledge
To achieve the diagnostic and therapeutic purposes for this paper, we note the need for a shared language. Without such a language, it is difficult to arrive at a consensual core body of knowledge or even to begin framing the issue of coding such a shared BoK for the discipline
From Where Have We Come 79
as a whole. Categorization schemes that make up the subject areas of IS (cf. Barki et al. 1988; Bacon and Fitzgerald 2001) are a useful start for developing a shared language for the field, but have not led to a discussion on how IS knowledge as a whole should be structured.
Moreover, most of the current research efforts have been devoted to knowledge that can serve a technical interest in prediction and control: ISD methods, tools, and other process knowledge such as database design, technology adoption, and so on. This type of knowledge is often widely dispersed (appearing in different disciplinary journals, e.g., management science, computer science, parts of applied psychology, etc.) and therefore not easily accessible. Furthermore, its relevance is often not seen because IS was unable to establish the kind of boundary spanning, social networks—what Klein and Lyytinen (2003) term ‘knowledge creation and transformation networks (KCTNs)’—that are possessed by other applied fields like medicine, law, and engineering. These disciplines have had a long history where members of their transformational network convert abstract research insights into understandable and action-oriented, practically relevant knowledge. The relevance of research results is not so much an attribute of their research papers’ wordings, but the product of the interaction occurring within a network of different agents and their motivation to transform knowledge similar to components in a food chain. Both the providers and the receivers must expend effort to communicate on the one hand and understand (interpret) on the other. To improve the relevance of its products, IS as a field must invest in establishing the appropriate knowledge creation and transformation networks. We shall return to this idea later in the paper.
In this paper, we shall propose a high-level classification scheme that includes practical, action oriented ‘applicative knowledge.’ We trust that our proposal will not endanger the currently very fertile pluralism that exists in the field. Our proposal for such a body of knowledge needs to be seen as merely a ‘trial balloon’ and not a concrete object. It is a ‘first cut’ to illustrate what we mean. It will hopefully lead to serious follow up research.
2.2 The Information Systems Field: A Reflective
History
The field of Management Information Systems (or simply Information Systems, as is now more commonly used) has been around since the 1960s and has been evolving ever since. It formed from the nexus of computer science, management and organization theory, operations research, and accounting (Davis and Olson 1985, pp. 13–14). Each of these areas or disciplines brought a unique perspective to the
80 Crisis in the IS Field?
application of computers to organizations, but each was also far broader in orientation. None focused specifically on the application of computers in organizations.
2.2.1 The emergence of theory
As computers began to be successfully applied to business problems in the late 1950s and early 1960s, interest grew in the development of ‘theory’ to support continued success. Blumenthal (1969) proposed what might have been the first comprehensive attempt at the development of a MIS (Management Information Systems) theory in his landmark book Management Information Systems: A Framework for Planning and Development. The author claimed his book ‘is the long awaited intelligent, scientific approach to determining an organization’s information needs and developing the kind of system that is responsive to sound decision making.’
Along the same timeframe as Blumenthal, Börje Langefors in Sweden was developing his own thoughts about IS. This culminated in Langefors’ (1973) seminal work Theoretical Analysis of Information Systems, which provided ‘a formal theory [of information systems].’ Langefors drew on systems theory noting: ‘. . . we try to support statements by drawing analogies from other systems theories, for which precise solutions to the specific problems or techniques for solutions, have already been devised (p. 17).’ Dickson (1968) proclaimed a dawn of a new era. He wrote: ‘A new academic discipline, “management information-decision systems” is emerging to integrate these techniques and to provide the analytical frames of reference and the methodologies necessary to meet the new management requisites (p. 17).’ Other authors rallied around the growth of MIS ‘theory’ and saw the development of total information systems solutions for organizations (cf. Young 1968; Zani 1970; Gorry and Scott-Morton 1971). But not all were enamored by the emergence of total information systems solutions (cf. Ackoff 1967; Dearden 1966, 1972). A debate ensued in the early years of the field about the efficacy of MIS (cf. Rapport 1968; Emery and Sprague 1972). Nevertheless, the field grew and flourished as discussed in Dickson’s (1981) thoughtful history of the field. There he noted that the ‘genesis’ of the IS concept could be linked to decision making and ‘viewing the management process as a cybernetic control system within the organization, relying heavily upon the computer as the control mechanism (p. 6).’
Keen’s (1987) articulation of the field’s mission reaches beyond Dickson’s focus on decision making and cybernetic control within organizations:
From Where Have We Come 81
The mission of Information Systems research is to study the effective design, delivery, use and impact of information technologies in organizations and society. The term ‘effective’ seems key. Surely the IS community is explicitly concerned with improving the craft of design and the practice of management in the widest sense of both those terms. Similarly, it looks at information technologies in their context of real people in real organizations in a real society. (pg. 3)
Because of its roots in multiple disciplines, such as computer science, management, and systems theory, it is hardly surprising that the field of IS cast a wide net when defining its boundaries, sweeping in many themes and areas. Nor is it surprising that there is considerable disagreement about what the field actually includes and does not include, and what its core features are. Mason and Mitroff (1973), for example, in their classic framework of IS, characterize the core components to be: psychological type (of the user), class of problems to be solved, organizational context, method of evidence generation and guarantor of evidence, and mode of presentation of the output. Ives et al. (1980) define IS in terms of five environments (external, organization, user, IS development and IS operations), three processes (user, IS development and IS operations), and an information subsystem. Lyytinen (1987) divides the field into nine components: the information system itself, IS operations environment, IS development environment, user environment, organizational environment, external environment, use process, development process, and operations process.
Swanson and Ramiller (1993) discuss the field in terms of the broad areas on which people write papers: computer-supported cooperative work, information and interface, decision support and knowledge-based systems, systems projects, evaluation and control, users, economics and strategy, impact, and IS research. Others have used co-citation analyses to identify intellectual subfields upon which IS draws (cf. Culnan 1986, 1987; Culnan and Swanson 1986; Cheon et al. 1992). Culnan (1986) for instance, noted the existence of three categories of ‘referents’ upon which IS draws: fundamental theory (e.g., systems science); related applied disciplines (e.g., accounting, computer science, finance, management, and operations research); and underlying disciplines (anthropology, political science, psychology, sociology). Keen (1987) categorized the field in terms of the problem areas each historical era chose to focus on. For example, in the early 1970s the focus was on ‘managing systems development, design methodologies, economics and computers.’ In the mid 1970s the focus changed to ‘decision support, managing organizational change, and implementation.’ In the early 1980s the focus was on ‘productivity tools, data base management, personal computing, organizational impacts of
82 Crisis in the IS Field?
technology, and office technology.’ And in the mid 1980s, it changed to ‘telecommunications, competitive implications of information technology, expert systems, impact of IT on the nature of work (pg.1).’
2.2.2 The beginnings of institutional infrastructures and diversity
The growth of the IS field over the past three decades has manifested itself in three ways. First, as the field has grown, new specialties and research communities have emerged, and the level of research has increased dramatically. Second, new journals, new conferences, new departments, and new IS programs are indicative of the dramatic growth of the field. We have witnessed the generation of a wealth of literature in information systems. Third, as the field moved into the nineties, this literature could be characterized as diverse and pluralistic. This is manifested in diversity of problems addressed, diversity of theoretical foundations and ‘referents’, and diversity of research methodologies (Benbasat and Weber 1996).11
Regardless of whether diversity is considered a blessing (e.g., Robey 1996) or a curse (e.g., Benbasat and Weber 1996), it is widely accepted as a defining characteristic of the field (Cooper 1988; Banville and Landry 1989; Alavi et al. 1989; Keen 1991; Orlikowski and Baroudi 1991; Swanson and Ramiller 1993; Markus 1997; Mingers and Stowell 1997; Mathiassen 1998; Benbasat and Zmud 2003; King and Lyytinen 2003).
In view of the rich diversity, it is not surprising that no unifying perspective on the nature of IS and IS research has gained widespread acceptance. The abundance of different schools of thought in the field is suggestive of its rich and diversified nature. The proliferation of different schools of thought in IS research, however, has its disadvantages as researchers in the different schools appear to work on disjoint or non-pertinent topics (Bjorn-Andersen 1984; Huber 1983) without much cross-communication. This has raised the concern that IS research does not contribute to a cumulative research tradition (Keen 1980), thereby throwing into question the value of IS research. To answer this challenge, several attempts have been made to find a common conceptual platform (paradigm) on which to ground, build and organize IS research (Mason and Mitroff 1973; Ives et al. 1980; Weber 1987; Farhoomand 1987; Ein-Dor and Segev 1981; Wand and Weber 1990; Hirschheim et al. 1996).
2.2.3 From diversity to the beginnings of fragmentation
However, Banville and Landry (1989) shed doubt on the possibility of achieving such a common conceptual platform. Through a sociological analysis of dependencies among researchers in IS, they concluded that
From Where Have We Come 83
the field is more a ‘fragmented adhocracy’ than a unified discipline. Why a fragmented adhocracy? Because in order to work in IS one does not need a strong consensus of one’s colleagues on the significance and importance of the problem to be addressed in research as long as there exists some outside community for support. Nor are there widely accepted, legitimized results or procedures on which one must build ‘in order to construct knowledge claims which are regarded as competent and useful contributions (Banville and Landry 1989, p. 54).’ In addition, there exists high task uncertainty in IS research, because problem formulations are unstable, research priorities vary among different research communities, and there is little control over research goals by a professional leadership establishment (such as bars or licensing boards for physicians and engineers). For example, IS research groups may choose to define projects that do not follow the familiar patterns of engineering or empirical social science. There exists considerable local autonomy to formulate research problems and standards for conducting research and evaluate research results. In fact, this has been a matter of lively debate for many years (Ives et al. 1980; Keen 1980; Mumford etal. 1985; Culnan 1986, 1987; Farhoomand 1987; Cash and Lawrence 1989; Benbasat 1989; Kraemer 1991; Nissen et al. 1991; Backhouse et al. 1991; Landry and Banville 1992; Galliers 1992; Hirschheim et al. 1995; Lee et al. 1997; Mingers and Stowell 1997; Checkland and Howell 1998; Currie and Galliers 1999; Klein and Myers 1999).
2.3 The State of IS Today
Today, we believe that fragmentation is a root cause of the field’s potential crisis. Whereas Banville and Landry (1989) described the field as a fragmented adhocracy, they did not explain why this condition arose nor whether this was a problem. Indeed, they believed it to be a strength.
In contrast, we believe the fragmentation is evidence of a structural problem for the field that portends a crisis. To be clear, we do not equate fragmentation with pluralism. Pluralism—for us—relates to diversity of ideas, perspectives, research approaches, paradigms, etc. But there is at least some underlying core set of knowledge or beliefs that all in the field share. There is a sense of shared belonging and empathy to others in the field. This is different from fragmentation, where there is insufficient (insignificant) communication between the different communities such that no core knowledge set exists. Individuals work in their own sub-communities without reference to other sub-communities. For us, this is a serious concern and in this paper, we shall explain why we feel this way. To do so requires us to